Electrochemical performance and failure mode analyses
The addition of dopamine to heparin was accomplished via the N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC) coupling reaction, as shown in Fig. 1. To prevent the oxidation of dopamine, the pH was maintained at 6.0 using a phosphate buffer solution (PBS) while the reaction proceeded.
A structural analysis of the dopamine-functionalized heparin (dopamine-heparin, 1) was carried out by performing UV-Vis and FT-IR spectroscopy investigations. The UV-Vis spectrum of 1 (Fig. 2a) showed a characteristic absorption peak at a wavelength of about 280 nm, confirming that the catechol moiety of dopamine was successfully grafted onto heparin32,34. In addition, the absence of any additional peaks at wavelengths greater than 300 nm indicated that no undesired oxidation of dopamine occurred32. The amount of dopamine incorporated was measured to be 11 wt.% of the total weight of dopamine-heparin 1, as determined by carrying out a UV-Vis spectroscopic quantitative analysis using dopamine standard solutions (see Supplementary Fig. S1)32.
Further analyses of 1 using FT-IR spectroscopic investigations confirmed the structure (Fig. 2b): the C = O stretching peak of heparin at 1620 cm−1 shifted to 1640 cm−1 after conjugation of dopamine due to the formation of amide bonding. In addition, a peak corresponding to the secondary amide NH bending appeared at 1560 cm−1 32,35, indicating that an amide bond successfully formed between the dopamine and heparin. Furthermore, a new peak corresponding to the aromatic OH stretching of the catechol moiety appeared at 1376 cm−1.
Preparation of electrodes and analyses of the mechanical properties
Electrodes were fabricated using a composite of SiOx and graphite as the active material, Super-P as the conducting agent, and CMC/SBR, heparin/CMC/SBR or dopamine-heparin/CMC/SBR as the polymer binder in a ratio of active material: conducting agent: binder of 80:10:10 by weight. The weight ratio of CMC to SBR was 1:2, and 10 wt.% of heparin or 1 relative to CMC was added to CMC/SBR for the fabrication of the heparin/CMC/SBR- or dopamine-heparin/CMC/SBR-based electrode, respectively. The results were compared with those obtained using the electrode containing only CMC/SBR as the polymer binder.
The binding affinities for the three electrodes (CMC/SBR, heparin/CMC/SBR and dopamine-heparin/CMC/SBR) were then investigated by carrying out 180-degree peel tests (Fig. 3a). In the case of the SiOx/graphite electrode composed of the pristine CMC/SBR polymer binder, an average binding affinity of 0.757 N was obtained, while this value increased to 0.907 N when heparin was added to CMC/SBR (heparin/CMC/SBR electrode). A dramatic increase in the binding affinity was further obtained for the electrode with the CMC/SBR binder containing dopamine-heparin (dopamine-heparin/CMC/SBR), with a measured value of 1.287 N. These results suggested that the addition of heparin to CMC/SBR mechanically strengthened the electrode due to the formation of physical interactions between CMC and heparin, and that the durability of the electrode in the presence of mechanical stress was further increased by the strong adhesion resulting from the inclusion of dopamine in the dopamine-grafted heparin (dopamine-heparin, 1)27,34.
We further used an optical microscope to examine the surface of the tape detached from the electrode after the peel test to determine whether the addition of dopamine-heparin (to form a dopamine-heparin/CMC/SBR binder) reduced the amount of Si exfoliated from the electrode. As shown in Fig. 3b, the amount of the electrode material detached from the current collector was inversely related to the adhesive force generated by the binder in Fig. 3a28. The tapes pulled off the pristine CMC/SBR-based electrode were stained with a lot of slurry and transmitted little light, indicating that much of the Si material in these cases became detached from the electrode and transferred to the tape. In contrast, the tapes pulled off the heparin/CMC/SBR-based electrode were stained with only a very small amount of slurry, and were observed to transmit a considerable amount of light, indicative of the better ability of heparin/CMC/SBR than of pristine CMC/SBR to prevent material from being detached from the electrode and consistent with the stronger adhesion force of the heparin/CMC/SBR binder based on the peel test. And this trend continued with the inclusion of dopamine: tapes pulled off the dopamine-heparin/CMC/SBR-based electrode, showed the least amount of slurry and transmitted the most light, consistent with the dopamine-heparin/CMC/SBR binder having displayed an adhesion force greater than those of the other binders.
After investigating the mechanical properties of the electrodes containing the heparin/CMC/SBR and dopamine-heparin/CMC/SBR binders, we then assessed the electrochemical performances of their corresponding cells. We performed the electrochemical evaluations at a low rate (specifically with a discharge rate of 0.2 C and charge rate of 0.5 C, where 1 C equals 450 mA g−1), and compared the results with the cell performance of the pristine CMC/SBR-based electrodes (see Fig. 4).
The voltage profiles of the prepared cells during the 1st cycle showed similar trends irrespective of the binder (Fig. 4a), suggesting that the addition of heparin or dopamine-heparin (1) did not cause any particular side reactions. In addition, the voltage profiles of both the heparin- and dopamine-heparin-based electrodes showed lower resistance during lithiation than did the electrode prepared from CMC/SBR. The enhanced physical properties of the heparin and dopamine-heparin binder systems have been suggested to result in less polarization due to the reduction of the contact resistance of the electrode36. Having the highest mechanical strength, the dopamine-heparin-based electrode indeed showed the lowest resistance.
In addition, the specific discharge capacities of the cells prepared using the three binders were also determined in order to assess the effect of the binder on cyclability (Fig. 4b). The specific capacity of the CMC/SBR-based electrode started to deteriorate rapidly after 50 cycles, and at 100 cycles showed a value of 286 mAh g−1 and a retention of 62.3% of the original specific capacity. After about 100 cycles, this electrode showed particularly poor cycling stability. This poor stability was attributed to the CMC/SBR binder, with its poor adhesion properties, no longer being able to accommodate the repeated changes in the volume of the silicon electrode. The cell performances of the electrode made of the heparin/CMC/SBR binder and that made with the dopamine-heparin/CMC/SBR binder, however, were much better than the cell performance of the electrode made with the pristine CMC/SBR binder. The superior performance of the former two cells was attributed to their enhanced physical properties. A specific capacity of 380 mAh g−1, representing a retention of 81.3% of the original capacity, was observed at 100 cycles, and a specific capacity of 268 mAh g−1 and retention of 57.3%, were observed at 150 cycles for the electrode including heparin/CMC/SBR. The dopamine-heparin/CMC/SBR-based binder system showed the best cell performance. Although the specific capacity and retention values of the dopamine-heparin/CMC/SBR-based binder system at 100 cycles, with values of 378 mAh g−1 and 81.3%, respectively, were similar to those of the heparin/CMC/SBR-based system, a much greater reversible capacity of 343 mAh g−1 and retention of 73.5% were obtained at 150 cycles for the dopamine-heparin/CMC/SBR-based system than for the heparin/CMC/SBR-based system. Since the three cells were subjected to the same conditions, the superior specific capacity and percent retention of original specific capacity up to the 150 cycles for the cell having the dopamine-heparin/CMC/SBR binder were therefore attributed mainly to the superior adhesion properties of this binder material.
The dopamine-heparin-CMC/SBR electrode also displayed excellent Coulombic efficiency (Fig. 4c). It displayed an initial Coulombic efficiency (after the formation cycles) of 97.4%, a value of 98.9–99.5% at the 20th cycle, and very high value of 99.7% at the 150th cycle. These values were higher than those of the CMC/SBR- and heparin/CMC/SBR-based electrodes.
We also performed rate capability tests for the three types of electrode (Fig. 4d). Their 0.1 C-capacities were similar to those in the formation stage of the cycling performance experiment. The CMC/SBR-based electrode (having the worst mechanical properties) showed a much larger decay of specific capacity as the C rate was increased; the performance of this electrode indicated that it would be difficult to use above 15 C. The heparin-based electrode, in contrast, exhibited a stable capacity up to 3 C. Although a constant capacity loss occurred when the rate was increased to above 3 C, the cell showed stable operation up to 20 C. The dopamine-heparin electrode showed stable performance up to 5 C, and showed a reasonably good cell performance of about 200 mAh g−1 even at 20 C. We attributed the enhanced rate properties of the heparin and dopamine-heparin electrodes to the improved physical properties of these binders, which helped to prevent the conducting agents and active materials from detaching from the electrodes even at high C rates. In addition, the functional groups in the heparin structure were thought to further provide a lithium transfer pathway capable of conducting Li+ ions28,37,38,39,40,41.
Electrochemical impedance spectra (EIS) were also acquired for the electrodes with and without the dopamine-heparin, and the overall resistance on the electrode surface was found to be smaller when the dopamine-heparin was included (Fig. 5). Since there were no other factors besides the addition of dopamine-heparin, this result was thought to be due to the formation of a solid electrolyte interphase (SEI) layer. Surface analyses using SEM and XPS further confirmed these results (see below).
Finally, a full cell, having a LiNi0.6Co0.2Mn0.2O2 positive electrode and the composite negative electrode containing the dopamine-heparin/CMC/SBR binder, was tested to evaluate the binder in a practical battery application (Fig. 6). To ensure reliable performance, one formation cycle was performed after the fabrication of the cell, followed by charging and discharging at 0.5 C between 3.0 and 4.1 V. Even though the initial high specific charging capacity of the full cell (Fig. S2), the obtained capacities at subsequent cycles suddenly fall because of the initial irreversible reaction of the negative electrode limiting the number of reversible lithium ions and the high kinetic resistance from the high amount of the active material causing high polarization. On the other side, an excellent cell performance, with retention of 92% of the initial performance at the 100th cycle and of 89% of the initial performance at the 150th cycle, was obtained, strongly suggesting that the proposed binder system has potential for use in practical Si-based lithium-ion batteries. In addition, the voltage profiles of the full cell (Fig. S2) showed 66.5% Coulombic efficiency at the pre-cycling step, which can be attributed to irreversible SiOx conversion and SEI formation during pre-cycling. After the pre-cycling, however, the system required only two cycles to show >99% Coulombic efficiency, which was maintained for more than 100 cycles (Fig. 6).
The CMC/SBR-based and dopamine-heparin/CMC/SBR-based electrodes were visualized before and after cycling using scanning electron microscopy (SEM), as shown in Fig. 7, in order to determine the effects of the binder material on the morphology of the electrode surface. The two electrodes showed similar surface morphologies immediately after fabrication, but quite different morphologies after 150 cycles, at which point the surface of the pristine CMC/SBR electrode was relatively thickly and unevenly covered by an SEI layer, while the surface of the dopamine-heparin/CMC/SBR-based electrode maintained a relatively clean, uniform and porous appearance.
Energy dispersive spectrometry (EDS) analyses of the prepared electrodes (based on analyses of mapped images of oxygen atoms, which originated from SiOx and the polymer binder, and of silicon atoms, which originated from SiOx) showed that the materials comprising the electrodes were well dispersed for both electrodes irrespective of the addition of the dopamine-heparin to the CMC/SBR (Fig. S3). However, the EDS results showed that the portion of the weight of the CMC/SBR electrode due to oxygen was about 22% greater after cycling than before cycling, whereas the portion of the weight of the dopamine-heparin-based electrode due to oxygen was only about 20% greater after cycling (Fig. S4). Since changes in oxygen content as a result of cycling have been suggested to originate from the formation of the SEI layer and/or the decomposition of electrolytes, these results might indicate that less electrolyte decomposed when dopamine-heparin was present.
We also carried out XPS analyses of the surfaces of the CMC/SBR and dopamine-heparin/CMC/SBR electrodes. For the CMC/SBR electrode, the peak intensity corresponding to C1s was much lower after 150 cycles than before cycling, whereas that for O1s significantly increased, confirming the formation of a thick SEI layer caused by the side reactions of the carbonate-based electrolytes (Fig. S5). In contrast, in the case of dopamine-heparin/CMC/SBR electrode, the changes in the peak intensities of the C1s and O1s peaks were not large even after 150 cycles, suggesting that there was relatively little SEI formation38,42. The XPS results were found to be in good agreement with those of the SEM analyses.